The Stellic Chatbot is a web application that assists users at Case Western Reserve University with academic scheduling. 
    It manages a student’s current academic process, degree requirements, courses taken, and classes offered at CWRU during the specified semesters. 
    The chatbot also provides AI functionality, allowing users to interact with it and obtain necessary information efficiently and effectively.

Install the required python packages:

pip install -r requirements.txt

1. Run the software on a website
2. Login using your name, Case ID, and select your major with the dropdown menu
3. After login, you can

• Access the chatbot with the chat functionality on the top left corner of the software and ask any question regarding major requirements and courses
• Use the search bar to search for courses

1. Run the software on a website
2. Login using your name and Case ID
3. After login, you can

• Access the chatbot with the chat functionality on the top left corner of the software and ask any question regarding major requirements and courses
• Create and add courses into the database

• AI/ - AI utility scripts
• scrape/ - Scripts related to scraping
• static/ - Static assets for the web (JS)
• templates/ - HTML templates for frontend
• tests/ - Testing files

• Backend
  • python
  • Flask
• Frontend
  • HTML
  • JavaScript
• Database
  • MongoDB
• AI Engine
  • Llama with Groq (RAG)
  • Sentence Transformers (Embeddings)
• Scraping
  • Selenium

Aidan Bugayong:

• RAG chatbot
• Frontend/Flask

Brandon Lee:

• Course/Degree Scraping
• Course Search

Natalie Lee:

• Unit Testing
• Frontend/Flask

Sam Lin:

• Database Management
• Admin Account

• Time Constraints for Scheduling Functionality
  • We originally planned to implement a course scheduling feature. However, due to limited development time and the complexity of scraping reliable scheduling data, we had to deprioritize this functionality.
    → Future improvement: Integrating APIs (if available) or using static sample schedule data could help in future iterations.
• Difficulty Implementing Vector Search
  • Integrating a working vector search was technically challenging and more complex and took more time than we anticipated.
    → Future improvement: Start with simpler keyword search first, then progressively move to vector models to ensure a working prototype earlier.

• Documentation is found in "docs/build/index.html"

This project is licensed under the MIT License.

Click here to view the GitHub project